CK Logo
Bootcamps
Projects
Post main image

Classification of Race and Ethnicity

Training a Deep CNN to predict the Race of Celebrities

#1 Project of March

Using a Deep CONVNET to Build a Model for Classifying Different Races such as Mongoloid, Negroid and Caucasian This kernel uses a deep CONVNET that was trained on Google GPU to perform Race Classification on a zipped file containing faces of different races. Each of the image are either labelled as: Caucasian: includes people of American and European descent, also known as whites Mongoloid: includes people of Asian descent, especially Eastern Asian Negroid: includes people of African descent or black Americans The zip Dataset contains various images of faces of different races which was aggregated from https://www.shutterstock.com/ I used this to build an face image classifier using a tf.keras.Sequential.model and I also built a input data pipeline using tf.keras.preprocessing.image.ImageDataGenerator. This project workflow includes: - Loading the zipped dataset from my google drive - Examining and understanding the dataset - Building a Data Image input pipeline - Building a Deep CNN Architecture - Training a CNN model - Testing the model - Using the model for prediction on new data All these was done with tensorflow 2.x.

Screenshots
Project screenshot 1
Project screenshot 2
Project screenshot 3
13
103